Virtalis is one of the biggest authorities on VR in the engineering sector. Greg Corke sat down with the company�s MD and Technical Director to talk about virtual prototyping and the expansion of VR into the realms of CAE.

Much has changed in the world of Virtual Reality. Ten years ago you'd need to fork out the best part on £1million for a so called immersive VR environment but today you can get started for under £20,000. But without the expertise to balance and tune these systems you could be throwing a lot of money at nothing.

Virtalis has a wealth of experiences in delivering VR solutions to the engineering sector and services a huge portfolio of clients from automotive and aerospace right through to defence and education. Projects range from the digital mock-up of complex components to the design and prove out of entire manufacturing processes - before anything is built. In addition, VR is playing an increasingly important role in visualising CAE data and with a ready source of 3D CAD data the potential for marketing is huge. In terms of technology, the company has vast experience in 3D graphics, tracking systems, haptic (force feedback) devices, projectors and active and passive stereo.

MCAD was lucky to sit down with Andrew Connell, Technical Director, and David Cockburn-Price, MD. Anyone who has met the Virtalis guys before will know they are more than a little passionate about their technology - excitable is not the word! An education in the world of virtual prototyping, 3D graphics and the increasing importance of VR for visualising CAE data followed.

Greg Corke: Who are Virtalis' customers? In terms of Virtual Prototyping you tend to think very much of the major OEMs.

David Cockburn-Price: In the engineering world there are certainly the major OEMs at the high end who have their own specialisms, ideas and uses for VR. What we're aiming to do is a) target them with new ideas and technologies to help them broaden and deepen the take up in their organisations, and b) start to get that filtered down into their supply chain - because you don't have to spend half a million or a million pounds plus to get started with some multi PC, multi projector environment. You can get started for 15-20 grand for a small, single channel solution, for example, and you can still get a lot of the benefits that the OEMs have got, and you can play your part as part of the supply chain.

So we're really going for the engineering market - there's lots of CAD, the CAD products are getting better and better, the 3D elements of CAD are getting better and it's really a case of saying you've got this big extra step into virtual prototyping, and that's what we're all about. As a result, we're changing our message. We're saying it's 'not just a design prototype'. CAD is bought by the design department in any engineering company, and the number of times we go and see them building models or commissioning special 3ds Max models of their products to take out to media launches and product events and exhibitions etc is huge. And what we're saying is "hold on a minute, whatever you do in the design department you should be able to take out into manufacturing and have them build it efficiently using your actual design data, then take it into marketing and also use it for training." There's a full product lifecycle aspect to it.

There are plenty of metrics out there that show how powerful the systems can be. You're talking 1000s of percent return on investment. You've only got to save one big (physical) prototype on a 20, 30, 40 grand investment and there's 1000s of percent return for you in a very short space of time.

GC: What barriers are there to adoption?

Andrew Connell: The problem you have with some people is that they question whether they will make a saving. They say: what if I don't find a fault? They're assuming they won't make any errors in the design. The other problem you get is that people who have used technology in the past which has promised these savings, they say you told me when you started using CAD that you weren't going to need physical prototypes, but we did, then the reason you made physical prototypes was because you didn't have electronic data management, and so on.

There's nothing that completely removes all physical prototyping, it's just that you reduce it, and reduce it, and reduce it. But we've been hit by people overselling in the past about what things can do, and overselling what visualisation can do as well. Virtual Prototyping can't guarantee you'll completely remove prototypes, but for certain customers it can.

GC: Can you give an example of a customer who has completely eliminated physical prototyping?

AC: Leyland Trucks is still one of the only companies we know of in that vehicle space that have completely eliminated the prototype. And this is partly due to the nature of truck assembly and partly because of the volume of truck assembly.

For example, if you're in automotive and you're going to make 500,000 of one design you'll probably still make physical prototypes, just because the cost per customer is so low. If it costs you, say, £500,000 for a prototype, a basic styling model, that's spread across a pound a car.

A truck prototype may be cheaper to make. It might only cost £250,000, but you might only make 10,000 trucks, or even 1,000, so suddenly your cost per truck goes up hugely, so the urge to get rid of them (a physical prototype) increases. And as you get down to the more speciality trucks, where you might only make one a month, the cost per truck for a prototype would be so high that no-one would ever buy one.

GC: What role does virtual prototyping play in the global marketplace?

AC: A lot of projects are multi-national now - they've got multiple sites for design and manufacture. And getting all the different people together and shipping prototypes is impossible, so they're doing virtual design reviews. By going to a virtual prototype and installing VR systems at each site they've been able to cut down on lots of travel and decisions can be made faster. We've got a lot of clients who say the reason projects have gone wrong is that they have not got enough people talking about a project. The more people you can get talking about a product, the more it'll improve, and new ideas will appear. If design review becomes expensive. i.e. you've got to ship 30 people 200 miles, it's not going to happen, or you won't get all the right people there and mistakes will happen. One way of making this cheaper is with VR.

Now with many companies outsourcing manufacturing, you'll get the manufacturing team in one country, the design team in another and virtual design review becomes even more important.

For the designers, if you're going to get a physical prototype built it helps if you're next to the guys with the skills that can put it together for you. If they're off in South East Asia somewhere, then the prototyping costs go up because you've lost your pool of skilled partmakers and toolmakers to do it for you. So you've got to find other ways of prototyping. If the manufacturing is overseas then the prototyping needs to become virtual, or the prototype build will go overseas and bit by bit the design will go overseas and eventually all you'll end up doing in the UK is specifying and marketing. Virtual Prototyping is a really useful skill to help keep that design facility in the UK.

Once the part is built, in China, for example, they can laser scan it and send it back to the design team and in the VR room they can compare the CAD model with the scan of the built part.

GC: So ideally you'd need really powerful hardware at every site?

AC: To some degree. But it doesn't have to be earth shattering. Everyone can run at their own speed. You don't even have to load the same data at each site if you don't want to. You can be drawing different amounts of data according to how fast your computer is. There are a lot of tricks you can pull. In answer to your question, a typical workstation costing £3,000 is enough for doing this and once you've spent more than £5,000 you don't get much improvement at all unless your proposed data is enormous.